The main problem with this analysis (there are others) is the use in the MOND Toomre criterion of the value of mu(a/a0) in the midplane of the disk instead of just outside the disk. They do not do the proper analysis but do note
`An important comment: we should have actually used mu+, but it doesn't really change much.'

This is absurd because
1. They could easily do and present the results of the correct analysis.
2. It does make a big difference. Near the center the midplane acceleration, and so also mu, vanish at R=0. This means that their wrongly calculated Q goes to 0 there perforce, so you always get instability there by their wrong criterion. If you look at their Q plots for the two galaxies they analyze, sure enough they approach zero towards the center. If you use the correct Q value it should go to a constant near the center.

"Response to Milgrom's comments by Sanchez-Salcedo & Hidalgo-Gamez:
If you do not assume that \mu is approximately constant along the thickness of the disk (in the doubful attempt to solve the problem of stability), then you will not able to fit neither the rotation curves nor the scale heights of these galaxies. The reasons of our choice were discussed in our paper. Part of the success of MOND in dwarf galaxies is that it is being used ignoring the vertical acceleration. The inverted commas in Milgrom's comments do not correspond to literal rephrasing."

In the standard Newtonian plus dark matter case, in which all the dark matter is presumed to be in a spherical halo, these galaxies are too stable. This is a general property of dwarf and low surface brightness disk galaxies. Getting bars and spiral structure requires some disk self gravity, but the dominant halos of these types of galaxies should suppress the sorts of structures which are observed. I think this presents as much of a problem for dark matter as it may for MOND.

It is worth noting that so far (as of 8/99) we are largely ignorant of the velocity dispersion in z, and only have V(R), especially of LSBs & dwarfs. The stability properties of the disk depend on both. In the [currently] conventional dark matter picture, it is presumed that the dark matter halo is quasi-spherical, and the disk purely baryonic [stars + gas]. For very low surface brightness disks, the stars will support only a small vertical velocity dispersion (a few km/s) in a purely Newtonian system. The z velocity dispersion of such systems would be rather higher in MOND. If we insist on interpreting this in terms of dark matter, then we would need to put a lot of dark matter in the disk itself, or infer a very flattened halo. This has the effect of raising the disk surface mass density, making it rather unstable in the Newtonian case (as opposed to overly stable if all the dark matter is in a spherical halo).

Addendum, 2002 In paper [38], I had pointed out that if MOND were correct, analysis of the stability properties of LSB galaxies in conventional terms would lead to the inference of lots of dark matter in their disks. This has since come to pass (Fuchs 2002, paper [166]).

Return to Literature page